33° BOTANICAL GAZETTE [NOVEMBER 
remained and were seen even in the pollen tubes. The pollen 
of all the species studied contained them, but they were most 
abundant in A. Cornuti. The tests for calcium oxalate, car- 
bonate,. phosphate, and sulfate, and for silicon and starch failed 
to reveal their nature. They stain like chromatin. 
A stage like that in fig. 76 was frequently found, and seems 
to indicate the fragmentation of the tube nucleus, although the 
series about it was not close enough to justify a definite con- 
clusion. 
Germination always bursts the pollinium along its outer edge, 
at the point where the diameter is greatest (fig. 77). A cross 
section shows the pollinium wall to be thinnest there. Germi- 
nation of the microspores was observed in A. Cornuti, A. tuberosa, 
and A, incarnata. 
SUMMARY. 
The development of the sporangia in the Asclepiadaceae 
studied is the same in general as in other plants, while there are 
no indications of the phylogenetic history of the reduction in 
number. The primary sporogenous cells without further divi- 
sion become the pollen mother cells. The latter divide each 
into four with the usual phenomena accompanying tetrad divi- 
sion, but through mutual adjustment and the close adherence of 
the microspores the evidences of grouping are lost. 
I wish to acknowledge my indebtedness to Dr. John M. 
Coulter and Dr. Charles J. Chamberlain for their kindly direc- 
tion and valuable suggestions. 
HULL Botanica, LABORATORY, 
THE UNIVERSITY oF CHICAGO. 
Note.— Just as this paper goes to press, one by Professor Strasburger, Zinige 
Bemerkungen 2u Pollenbildung bei Asclepias (Ber. deutsch. bot. Gesells. 19 : 450-454- 
1901), dealing with Asclepias Cornuti and Cynanchum Vincetoxicum, has come to 
hand, confirming in a general way the results reported above. 
EXPLANATION OF PLATE XIII. 
The figures are reduced to one half their original size. The lenses used 
were Leitz objectives 3 and 7, and Zeiss *s oil immersion; oculars, Reichert 
